Evaluation of features for SVM-based classification of geometric primitives in point clouds

  • In the reverse engineering process one has to classify parts of point clouds with the correct type of geometric primitive. Features based on different geometric properties like point relations, normals, and curvature information can be used, to train classifiers like Support Vector Machines (SVM). These geometric features are estimated in the local neighborhood of a point of the point cloud. The multitude of different features makes an in-depth comparison necessary. In this work we evaluate 23 features for the classification of geometric primitives in point clouds. Their performance is evaluated on SVMs when used to classify geometric primitives in simulated and real laser scanned point clouds. We also introduce a normalization of point cloud density to improve classification generalization.

Export metadata

Additional Services

Share in Twitter Search Google Scholar
Metadaten
Author:Pascal Laube, Matthias O. FranzORCiDGND, Georg UmlaufORCiDGND
DOI:https://doi.org/10.23919/MVA.2017.7986776
ISBN:978-4-9011-2216-0
Parent Title (English):Proceedings of the Fifteenth IAPR International Conference on Machine Vision Applications (MVA), 8-12 May 2017, Nagoya, Japan
Document Type:Conference Proceeding
Language:English
Year of Publication:2017
Release Date:2019/05/29
First Page:59
Last Page:62
Note:
Volltextzugriff für Angehörige der Hochschule Konstanz möglich
Open Access?:Nein
Relevance:Keine peer reviewed Publikation (Wissenschaftlicher Artikel und Aufsatz, Proceeding, Artikel in Tagungsband)
Licence (English):License LogoLizenzbedingungen IEEE